People who know me well know that I’m particularly fickle when it comes to computer operating systems. I think this is probably a specific manifestation of OCD, but one that safe to tolerate as it has not actively reduced my ability to enjoy real life 
Recently my fascination with OS flavors has led me to start toying around with Windows 7, and my brief experience with it has led me to believe that the threat of upstart operating systems like Linux and the as-yet-unseen Google Chrome OS is only a small part of Microsoft’s current nightmare.
The migration of applications and services to the web (Google Docs, et al) is probably devaluing the Microsoft pantheon of applications, but not as widely as individual open-source contributors may think. Microsoft Office is the industry standard; it has been adopted by enterprise (read: organizations that pay actual money for actual software) and in many cases extended with custom-built VBA “glue” that would be more costly to replace than it is to maintain. You, the individual, have a choice–go download OpenOffice–but even if I were given the opportunity to change office platforms for my company tomorrow, I’d be hard-pressed to justify the expense (deployment and re-training efforts) of the mass migration to open source.
But that’s not really the point I’m trying to make here. What is far more interesting to me is that the past year has brought the world face to face with a fundamental change in the computer business, and it is kicking Microsoft’s ass.
I think this is a situation that is on par with what happened to Digital Equipment Corporation. Do you remember DEC? In the late 1980′s, it was the second-largest computer company in the world with one hundred thousand employees. For a lot of reasons, DEC didn’t make it into the new millennium, but primarily the problem was this: DEC was a really big company with a lot of momentum, not a lot of flexibility, and a particular sense of invincibility regarding its products. Personal computers (IBM PCs running DOS, which was written by that upstart company called Microsoft) arose as the direct competitor to DEC’s highly successful, proprietary and expensive VAX systems. PCs took off so fast that DEC more or less stood around in disbelief and then disintegrated. The interesting parts of DEC belong to a number of companies now, most notably HP.
DEC was in the hardware game, and yet it was a change in the hardware game that took the company down. Microsoft has suddenly found itself in an uncomfortably similar situation. I believe that Microsoft inadvertently got into the hardware game when it started its “Certified for Windows Vista” program. To qualify this: Microsoft has always been “in the hardware game” to the extent that Windows is inextricably linked to the x86 instruction set. (For you non-techies, that simply means that Windows won’t, for instance, run on your watch, or on your microwave–unless your watch or microwave has the kind of chip in it that people usually put in personal computers.) But the x86 processors are ubiquitous. On the other hand, at the time that Windows Vista was released, high-end video cards, systems with > 4GB RAM, and other trappings of the hot-rod end of the PC spectrum were not. Certified for Vista turned out to be a bit of a scam, and the the bottom line was this: if you wanted Vista, you needed some pretty serious power.
It is no secret that Vista was a commercial failure. Like companies that make cutting-edge video games, Microsoft was counting on Moore’s Law to provide the market with higher-end systems and in fairness, hardware manufacturers provided. But people weren’t buying those systems. And their current computers simply couldn’t run the new OS. Remember the enterprise sector? The large entities that pay real money for real software? The enterprise market basically patted Microsoft on the shoulder and said there was no way in hell they were going to wholesale replace every PC in their organization to run a new OS that added absolutely no value to their business. Ouch.
The worst part, though, is this fundamental shift in the market that I really don’t think Microsoft perceived as a threat. Moore’s Law suggests that we get a doubling of processor power every two years. If you are building an operating system based on the idea that by the time you get it to market, the average PC will be four times as powerful as when you started, you are probably not overly concerned about inefficiencies. Sure, it’s a little slow now, but when that 4x power CPU comes along, who’s going to notice?
Only, this year, in addition to the anticipated uber-CPUs that Microsoft was banking on, the market was flooded with a new class of processor–super-cheap, low-power CPUs that have about as much processing power as cutting-edge CPUs from six years ago. Do you remember what computing was like way back then? If you’ve mostly been using your computer to do word processing and web surfing, like the overwhelming majority of daily computer users, then you probably recall that your experience was pretty much the same as it is now. That’s because while requirements for bleeding-edge showcases like video games (and, you know, operating systems like Vista) have pressed onward, requirements for everyday applications like web browsers and Microsoft Office have basically not moved–and you have the enterprise sector to thank for that.
The result of this situation is that these cheap power-miserly CPUs can get paired up with comparatively large amounts of RAM (1 to 2GB is large, trust me) to deliver a daily computing experience that is just as good as the one you can get from a fresh-off-the-line quad-core powerhouse. So if you’re Microsoft, and your new operating system was banking on Moore’s law, things suddenly look bad for you.
To be fair, Microsoft is still competing quite well in the “netbook” market that has sprung up around these low-power systems. The problem is that they are doing it by extending support for the now-venerable Windows XP. Microsoft has officially stopped selling XP, but you can still get it brand new on netbooks and a handful of real desktops, as well. This means that Microsoft has to continue to support it, taking resources away from an operating system that they’ve recently dumped hundreds of millions of dollars into. Meanwhile, competitors to Windows XP abound, and the longer XP stays in the game, the more easily interchangeable it will become. I have my own misgivings about Linux, but I think this is fairly telling: beyond the cosmetic differences of Windows XP and Ubuntu (a popular distribution of Linux), the non-computer-geeky members of my household have no difficulty whatsoever finding their way around Ubuntu for accomplishing daily tasks.
So now we’re watching Microsoft’s new strategy in play. Don’t think that Windows 7 isn’t Vista. It is Vista, but Vista’s got such a bad rap that MS has to ditch the identity. The UI only is different enough to give a “brand new” first impression. Under the hood, Windows 7 is Vista, augmented with the kind of solution-wide backwards-compatibility that the enterprise market requires. And the advent of netbooks has set a highly visible and uncomfortably low bar for OS power that Windows 7 absolutely must meet to unseat XP.
There is no doubt in my mind that people have been pulling some long hours in Redmond. I’d be a little jumpy if I realized that I was about to go down the same way as the Goliath that I had brought down twenty years previously. But if the Windows 7 Beta that I have been playing around with is Microsoft’s best effort at reducing Vista to a retro-friendly flying gas can, then Microsoft isn’t out of trouble.
I’ve run three other operating systems on my MSI Wind U100 (it’s a netbook), and every one of them, including Windows XP, has run faster than Windows 7. When I moved Windows 7 to my fairly cutting-edge desktop system to continue the experiment, I found the slick UI and a reasonable stab at software backwards-compatibility. Unfortunately, and I think this is really going to hurt Microsoft the most–the backwards compatibility must be initiated by the user in many cases. If I want to install software originally coded for XP I have to watch the installer fail in a completely bizarre way, potentially leaving cruft scattered throughout my system. Then I have to go back and manually tell it that this app was written for XP and please do the right thing to make it work here.
I said before that the enterprise market is what Microsoft cares about the most, so we’ll see what the enterprise market has to say about Windows 7. But if, as I predict, the low-power CPUs will be heavily influencing the success of this OS, then Microsoft is looking a lot like DEC right now, and they know it.
Whereas I agree with a lot that you’ve said, I still think Microsoft has got some life in them still. Mostly, they’ve clinched the enterprise and business markets. Once another contender starts working their way into this area, then I think they’ll be in much bigger trouble.
Who is that going to be ?
Like you mentioned above, there is a lot of overhead moving to another platform, or even another application (like Office) for businesses. It’s not going to be worth it to a lot of them.
Linux is still not ready for most business markets. Don’t get me wrong, it is being used, but it’s compatibility to existing Microsoft infrastructures is poor, and until that improves, most folks will pass it over.
Apple ? As much as I like their new OS and such (Yay pretty GUIs on Linux), they also have a fair amount of compatibility issues with windows. Also, in terms of business sense, they are more expensive to purchase and maintain as many IT folks don’t know it as well as their Microsoft counterparts.
I’d more likely compare Microsoft’s issues to those of Apple after their celebrated desktop breakthrough in the 80′s. Back then, the apple’s were great machines to work on, for productivity, as well as graphics arts, etc. Once Microsoft came along and put it’s Windows OS on the generic PCs, folks began to have some options for how much they wanted to spend, and purchase third party upgrades, etc. Apples usage started to decline and they came pretty close to going away until the iPod saved them. Now they’ve hit a resurgence, but there were many years of pain in there.
I think Microsoft as the giant it is will be going through a painful period of time as well, just like Apple. The “cloud” computing phenomena is taking off and folks are more willing to trust their data on public servers for collaboration and ease of access.
Once upon a time we went from the dumb client and server model, to having everything on our own desktops, and now we are heading back the old way. Netbooks are popular mostly because of their easy of portability, but what do folks really use machines for now a days ? It’s all web based. 99% of what the typical user wants to do can be done via the web. Why would they need stupid powerful computers ? Those will become more specialty models in the future where developers, games, etc will need them.
Once Microsoft learns that it needs to start selling services, or find their way into the cloud computing arena, they will be fine. Though I believe they will hold out as long as they can.
(I’m still interested to see how much folks will really put themselves into cloud computing once the first disaster of personal data being distributed finally happens. We’ve seen some small things up til now, but I’m sure something bigger will hit . . . will their confidence in such a thing make them go back to the old ways?)
As far as Windows 7 goes, I’ve had a lot of success with the recent releases. I’ve been able to install all of my software, it’s fairly speedy, and I haven’t seen any real issues. I’ve installed it on a few machines about 5 to 6 yrs old and aside from the Windows Media Center, it’s been running fine. Does that mean it is the next best thing ? No. It took XP about 3 or 4 years of updates to make it what it was at the end. My initial impressions of 7 are much better then Vista, but I do believe there is still work to be done. It’ll probably be the next OS I go to from XP. (Of course, I still have two flavors of linux floating around as well as playing with some of the Android stuffs).
Just opinions, but I’m keen on seeing how things turn out. It’s going to be an interesting couple of years.
Awesome feedback and great points! Thank you!
Don’t get me wrong; I think Windows 7 runs well on many systems–but given that 1) we are moving towards cloud computing and that 2) you can participate in the cloud with some pretty low-end hardware–I think that Windows 7 is not svelte enough to compete on those low end systems and that may be a lose.
Think of it this way: let’s say I wanted to start a business tomorrow with satellite offices all over the country. Now lets assume that the only requirement of the computers is that they need to be able to run Word, Excel, Outlook and IE. Right now I can buy Atom-based net tops for $220 a pop and manage them all from a single Windows Server machine.
Based on my experience running Windows 7 on a nearly identically spec’ed netbook, I am certainly not going to pay for the upgrade to put on those systems. In addition to being expensive (probably more than doubling my cost per PC), those system will run slower–that’s an enterprise wide decrease in productivity.
That’s what I’m really getting at. Sure–the OS works–but the functionality of the OS is not significant anymore. Even for my locally installed applications I’m getting better performance out of XP on that platform.
Interesting points and I agree with most of them. I think that Microsoft has banked on their name for a long time and, as you mentioned, largely depend on enterprises to deliver a huge chunk of their revenue. We could banter back and forth about UI supremacy, but I believe that it is ultimately about what you am do and how easily you can do it. Companies are looking to virtualize as much as possible to reduce power and hardware footprint and Linux is a great way to do that as it is light weight in comparison to Windows. Bot only that, but it’s easier to install and remove software (no registry).
The consumer market really cares about cost and, unfortunately, Apple remains a boutique brand. As long as they can retain that cache, they’ll do just fine. They have less than 10% of the market, yet have huge cash hordes. I’m not sure if the OS will become hidden to the users or if items will retain their architecture specific nature, but the web with faster bb speeds and cloud computing bring a lot of hope to the removal of software on the home pc (see Google Docs and Photoshop online).
Time will tell.
What kills me is that people aren’t even aware that Apple’s iWork is 100% totally compatible with Office, both on import and export. It costs under $100 for five licenses of Pages, Numbers, and Keynote (Word, Excel, PowerPoint, respectively).
It’s also easier to use and makes documents/spreadsheets/presentations that are much, much prettier. It even has online web-based collaboration tools in beta at iwork.com. It boggles the mind that people just aren’t interested in it, but I think it has to do with Apple’s lack of marketing for the product to anyone outside of an Apple store and the requirement of buying Apple hardware (though the cost barriers to entry there keep pushing downwards). I expect within another couple years Apple will try to make more of a push into home offices and small business with iWork; not that Microsoft ever cared about that user base – they’re all about enterprise, as you said.
Re: 100′s of millions of R&D – no kidding. I think MS R&D is roughly 10 B. a year. Re: Todd’s comment Apple – technically it’s FreeBSD underpinnings and not Linux.
Working in the XML / middleware arena, I was amazed at how deeply entrenched microsoft is. Now working on database technology – again Microsoft is very prominent with SQL Server. I don’t think you can look at a company that builds a full stack – from desktop OS and office applications, through the messaging layers, middleware stacks, all the way to world class database and back-office software through quite as narrow an aperture as this. They are a broad and deep technology company (for example their search technologies have just displaced Yahoo!) and their technology innovations force followers (see .NET clones on linux; web service stack integration in middleware; office interoperability per: Mike’s comment.)
I think the DEC analogy is well chosen – DEC was also a deep technology company (and I think a fair bit of their tech still lingers about re-branded and mutated in other companies). But I wonder if the “web-platform” as paradigm shift is less acidic to MS than personal computing was to DEC. MS has the resources to build web-scale services (again, the search thing and their partnership with yahoo as examples) and the web-as-platform is relatively OS independent. The web, at a consumer level, doesn’t seem to advantage Linux v. Windows or OS X v. windows. MS can control DRM and use its still dominant browser and OS market share as barriers to entry (still adobe flash sucks on my linux boxes to the point of being unusable; I don’t even try WMV or Quicktime there).
Computers are much more commoditized now than when DEC failed. The web-as-platform offers interoperability and features that to me lessens the end-user-value of most desktop platforms – but does this just make consumer-support, help, documentation all the more important? These are areas where Linux or other free systems will stuggle to catch up. (Though Canonicals efforts to organize major upstream linux software package release schedules is fun to dream about.)
DEC failed to follow the computing market – is MS failing at this as drastically? They took share from the game market with XBox; they will fight for share of streamed video and other entertainment markets (with the xbox platform); they will continue to gather vast revenues from Windows simply because no one else is ready to support the OEMs and low-level-consumers… and I think they work relentlessly at building profitable web products. Also MS seems to have transitioned from its founding leadership team to other capable leaders – something DEC didn’t really do, as far as I remember.
Thought provoking text. I’m not sure I’d bet this strongly against MS but it looks like it’s shaping up as a period of time that will react to the impact of internet SaaS on consumer behavior. I really wonder if Microsoft worries more about netbooks (or IMO more likely smart phones where they are also failing) or more about {Flickr, Evernote, Facebook} or {Amazon S3/EC2, Cloudera, RightScale} or {Google} or {NetFlix, Apple online media services}.